Dear SPMers,
I have encountered a problem while running a PPI analysis with the gPPI toolbox. I have run a PPI analysis for several different ROIs that were active in my task, defined the contrast of interest for each subject in the first level analysis and used the contrast images for second level analysis. Everything has run fine except for one PPI analysis for one region, for which the t-map in the second level (one-sample t test) showed strange results, which were evident when thresholding the t-map at p=1 (see attached image).
A closer look to the data revealed that the problem was caused by one subject, which showed unusually high values in its contrast image (ranging from approx. 200-300 to 2000, both positive and negative values, while the other subjects showed much lower values). When I excluded this one subject from the analysis results were normal. However, I have been unable to find out the reason for this extreme values and why they have appeared only for the PPI analysis for one ROI, but not for the other ROIs or the initial analysis of task main effects. The contrast vector and onsets for this subject seem normal and there is no excessive motion during the task.
Has anyone encountered a similar problem? Should I exclude this subject from my analyses, even from those that were not problematic?
Any insight will be very much appreciated.
|